CASE SELECTION; AND SAMPLING Susan Kilonzo Case studies • A case could be anything (event, process, activity, actors…). • However, a case can analytically be separated from its environment. • It is therefore a spatially bounded phenomenon observable at a given time. • Case Study is an intensive analysis (qual or quant) of a single unit (case) or a small number of units (cases) • A detailed analysis of the relationships between the contextual factors and an natural occurrence e.g. (GBV and natural disasters?) 3 What is a case? • A case could be anything (event, process, activity, actors…). 4 Determining a Case in a Hospital Study- Several units In defining your case… • Determine : • What do I study? – region, an individual, an organisation, a process, a system, activity? • What are your decision points in defining cases (analytic gains and losses of these boundaries)- Success story of Botswana • Determine what the case is not Bounding: • A case has two boundaries which are to be well defined: space and time 6 Determining a Case • The research question determines cases we select. • Example 1: How has Devolution Influenced Access to Referral Maternal Healthcare in Western Kenya? (Kilonzo et. al) • Perceptions from users and providers • Example 2: How has devolution improved access to referral maternal healthcare in Kisumu and Uasin Gishu Counties in Kenya? Kilonzo et. al. 7 More examples • How does land consolidation affect agricultural production in Amhara and Afar regions of Ethiopia? • How does local leadership influence different election outcomes in Nigeria? • To what extent do the interventions on FGMs vary across different regions in Northern Sudan? • How has the implementation of free day secondary education improved student retention in Kenya? 8 Determining a Case • Consider your decision points in defining cases • Ask yourself: • do I want to “analyze” the individual? • Do I want to “analyze” a program? • Do I want to “analyze” the process? • Do I want to “analyze” the difference between organizations? • In comparative cases, ask yourself, what is 9 Separating/Bounding a case • Once you determine what your case is, also consider what it is NOT. • A case can analytically be separated (bounded) from its environment. 10 Strategies for bounding a case: • Time and space: How has Devolution (introduced in 2013) in the health sector Improved Access to Referral Maternal Healthcare in Kisumu and Uasin Gishu Counties? • Time and activity: How has devolution improved access to referral maternal healthcare for women working in the informal sector? • Time, definition and context How has devolution improved access to referral maternal healthcare for middle class women in rural areas? 11 Selecting cases • Considerations… • Purpose for which cases are selected/goal of researcher (causality; causal mechanism, description, exploration, comparison). • Availability of data But why not choose cases randomly to avoid biases? Case Selection (Gerring, 2007) 13 Determining how to do a case study 14 Determining paths in case studies • Individual case (interest is details of one case, not to compare across cases) • Comparative (differences and similarities) • Causality (causal chain/mechanism). Sampling What is Study Population? sampling? Sample? • The word ‘population’ is used to describe the target group. • A sample is a section of units from the wider population which will be engaged in the study. • Sampling is the process of identifying who you will aim to contact from that population. • A sampling frame is a list of members of a population from which members of a sample are then selected. Sampling Techniques • Sampling is categorised into three: • Probability sampling • Non-probability sampling • Mixed Method Sampling Probability Sampling Probability sampling allows all the elements of the target population stand a chance of being selected into the study sample. These are primarily used in quantitatively oriented studies and involve ‘‘selecting a relatively large number of units from a population. Types: • Random sampling (ballot & systematic) • Stratified Sampling • Cluster sampling • Multi-stage sampling (sampling across hierarchical levels) Non-Probability This is sometimes called purposive sampling. Types of Purposive Sampling • Sampling to Achieve Representativeness or Comparability • Sampling Special or Unique Cases • Sequential Sampling • Sampling Using Multiple Purposive Techniques Types of Convenience Sampling (Non-probability) • Captive Sample • Volunteer Sample Mixed-Method Sampling Procedures Basic Mixed Methods Sampling • Sequential Mixed Methods Sampling • Concurrent Mixed Methods Sampling • Multilevel Mixed Methods Sampling • Combination of Mixed Methods Sampling Strategies